Power Draw, In-Flight versus Bench

NateKendrick

Junior Member
One thing I've noticed in a lot of advice columns and articles is that most discussions of power draw are based on bench testing. Hook up the battery, ESC, motor, and prop, with a watt-meter in-line. Run it up, and check the power draw. Put it on a force balance and check the thrust.

This makes sense for quad-rotors, but in-flight, with forward speed, the power draw and thrust of an airplane prop will be much less than static.

http://www.apcprop.com/v/downloads/PERFILES_WEB/datalist.asp

For example, from the APC web site, an APC 9x6E has the following performance data (condensed):
PROP RPM = 8000

V__ J___ Pe____ Ct____ Cp___ PWR_ Torque Thrust
(mph) (Adv Ratio) ____________(Hp)_ (In-Lbf) (Lbf)
02.1 0.03 0.0482 0.1006 0.0634 0.154 1.214 1.346
10.4 0.15 0.2251 0.0814 0.0549 0.134 1.052 1.089
20.7 0.30 0.4125 0.0644 0.0474 0.115 0.909 0.862
31.1 0.46 0.5614 0.0501 0.0407 0.099 0.779 0.670
39.3 0.58 0.6494 0.0374 0.0332 0.081 0.636 0.500
49.7 0.73 0.6884 0.0196 0.0208 0.050 0.398 0.262
60.1 0.88 -0.0003 0.0000 0.0053 0.013 0.101 0.000

Note that static it draws .154 HP (115 W) but at 40 mph it drops to 0.081 HP (60 W). This would mean about half the current draw. Thrust has dropped from 1.3 lbs to only .5 lbs. This plan will start like a rocket and climb fast, but have low top end speed.

An APC 8x8E on the other hand:
PROP RPM = 8000

V__ J___ Pe____ Ct____ Cp___ PWR_ Torque Thrust
(mph) (Adv Ratio) ____________(Hp)_ (In-Lbf) (Lbf)
02.6 0.04 0.0580 0.0958 0.0714 0.096 0.760 0.800
10.5 0.17 0.2157 0.0997 0.0799 0.108 0.849 0.832
21.0 0.35 0.4012 0.1031 0.0888 0.120 0.944 0.861
31.4 0.52 0.5503 0.0945 0.0891 0.120 0.947 0.789
39.3 0.65 0.6411 0.0813 0.0822 0.111 0.874 0.679
49.8 0.82 0.7353 0.0611 0.0682 0.092 0.725 0.510
60.2 0.99 0.7998 0.0385 0.0478 0.065 0.509 0.321
70.7 1.17 0.7669 0.0134 0.0204 0.027 0.217 0.112

Notice how the power draw is low at 2 mph (it's stalled), and peaks between 20 and 30 mph. At 40 mph, it has roughly the same efficiency (Pe), but it both draws more power (83 W) and provides more thrust (.68 lbs). At 50 mph it produces twice the thrust of the 9x6E.

The interesting thing to me is, say you're sizing your system, and you want to make sure you don't pull over your amp ratings. If you bench test, it might show too much. But unless you're doing 3D, you'll never see that kind of draw except for the split second you when you take off. Once the aircraft is moving, current load drops off quickly. You could conceivably use a lower spec (lighter) motor and never risk your system. And calculated flight times, even at full throttle, should be better. It's all well and good if you want safety margin, but if you want to maximize performance, it seems there is a lot of room for trimming specs.

This may be old news to old hands, but I wanted to open the discussion and see what I could learn. Any tips from more experienced builders?
 
Last edited:

Jaxx

Posted a thousand or more times
This is great information! Do you mind positing a link to that data? Thanks.
 

edfjockey

Junior Member
This data is also not completly accurate as these are wind tunnel tests with just a motor and prop mounted to a force gauge post. Yes it will give you a great idea on starting points and relative prop changes to draw and thrust. This was done in the 80's by airscrew for their props on glow and gas engines. What this doesn't take into account is the airplane design and drag that will keep the draw up just pulling the airframe through the air. Nice to see this type of chart again though. Keep 'em flying.
 

quorneng

Master member
As edfjockey points out you need to be careful with a such a table.
It refers to power absorbed at a constant prop RPM.
Any motor, electric in particular, slows down with load so as the prop unloads due to forward speed the motor will also speed up.
Yes the power will reduce but not quite as defined by a constant RPM table.
 

NateKendrick

Junior Member
quorneg and edfjockey, I think the data does account for the affects of inflow velocity. The data I excerpted is only a small portion of the data on the web site. The data on the web site is organized by prop, then rom, then tabulated by inflow velocity. For each, it provides torque, power, efficiency, and thrust. So you could interpolate between different rpm tables to account for the increase in rpm as the prop unloads. Note, the power is the power input to the prop, while power out would be thrust times velocity, which is given in the efficiency (Pe).

For example,
PROP RPM = 8000
V__ J___ Pe____ Ct____ Cp___ PWR_ Torque Thrust
mph AdvR _________________ Hp___ In-Lbf Lbf
43.5 0.64 0.6798 0.0305 0.0286 0.070 0.549 0.408
45.6 0.67 0.6890 0.0270 0.0261 0.064 0.501 0.361
47.6 0.70 0.6927 0.0233 0.0235 0.057 0.451 0.312

PROP RPM = 9000
V__ J___ Pe____ Ct____ Cp___ PWR_ Torque Thrust
mph AdvR _________________ Hp___ In-Lbf Lbf
51.2 0.67 0.6998 0.0270 0.0257 0.089 0.624 0.456
53.6 0.70 0.7034 0.0233 0.0232 0.080 0.561 0.395
55.9 0.73 0.6992 0.0196 0.0204 0.071 0.495 0.332

For a given airplane, drag is a combination of induced drag, which falls off with speed, and profile drag, which increases with speed. At the top end, profile drag dominates, and goes up as the square of the velocity. Say a plane's drag is 0.30 lbf at 48 mph, we could estimate it's drag at 55 mph as (0.30)/(48)^2*(55)^2 =.39 lbf.

With the prop above, it would have spare thrust of 0.01 lbf at 48 mph to accelerate up to about 54 mph. Power draw would increase from 0.057 hp (43 W) to 0.080 hp (60 W) at a higher rpm. If the motor is only 1000kV on a 3S, then it may not be able to go faster than 9,000 rpm because of the motor characteristics. And because you need more power at low flight velocity because of the increased torque and current draw, you'll have a large power margin at top speed. A smaller motor could reduce weight and increase performance by flying closer to its limits, but you'd run the risk of over-torquing it at low speeds (take-off, go-around, recovery...)

In a wind tunnel, you could always increase the inflow velocity up the point where it reaches pitch speed and the prop is completely unloaded (except for it's own profile drag). An actual airplane is going to find the balance between airplane thrust-required and the thrust limits of the motor/prop combo. My curiosity is that this will often be a motor rpm limit (based on kV, V, Rm, and I_0) instead of a current and power limit.

So when doing motor sizing for something other than 3D, can you trim your specs by sizing for stall speed? Say, pick a prop that will hit your power/current limit at 10 mph instead of static? Or even riskier, if you're going for top speed, size for something that would over-draw at low speed? As long as it has spare thrust at lower throttle, you could get slowly/safely wound up to full speed as the prop unloads. Telemetry would be essential, I think.